Research map / Label-efficient and robust learning / Recognition and long-tail
Decision Boundary-aware Generation for Long-tailed Learning
Jiacheng Yang, Ruichi Zhang, Chikai Shang, Mengke Li, Xinyi Shang, Junlong Gao, Yonggang Zhang, Yang Lu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add prototype-distance filtering to generator (ablation) | DBG · CIFAR100-LT (imbalance 100, ablation) | overall accuracy: 47.0 -> 50.1 | Improved |
| add confidence-credibility filtering to generator (ablation) | DBG · CIFAR100-LT (imbalance 100, ablation) | overall accuracy: 47.0 -> 47.1 | No clear effect |
| add confidence-credibility filtering on top of prototype-distance filtering (ablation) | DBG · CIFAR100-LT (imbalance 100, ablation) | overall accuracy: 50.1 -> 50.4 | Improved |
| add both filters to the boundary-aware generator (ablation) | DBG · CIFAR100-LT (imbalance 100, ablation) | tail accuracy: 23.2 -> 26.7 | Improved |
| add boundary-aware generator without filtering to CBDM* (ablation) | CBDM* · CIFAR100-LT (imbalance 100, ablation) | overall accuracy: 48.8 -> 47.0 | Hurt |
| add DBG samples to CBDM* | CBDM* · CIFAR100-LT (ResNet-32, imbalance 100) | overall accuracy: 48.81 -> 50.37 | Improved |
| add DBG samples to DiffuLT | DiffuLT · CIFAR100-LT (ResNet-32, imbalance 100) | overall accuracy: 49.72 -> 51.07 | Improved |
| add DBG samples to DiffuLT | DiffuLT · CIFAR100-LT (ViT-B/16, imbalance 200) | overall accuracy: 26.90 -> 30.42 | Improved |
| add DBG samples to CBDM | CBDM · CIFAR10-LT (ResNet-32) | average accuracy: down 0.13 average | Hurt |
| add DBG samples to CBDM | CBDM · CIFAR100-LT (ViT-B/16, imbalance 100) | tail accuracy: 16.30 -> 15.63 | Hurt |
| vary prototype filter ratios h and l (ablation) | DBG · CIFAR100-LT (ResNet-32, imbalance 100) | overall accuracy: 49.3 - 50.4 | No clear effect |
| remove lower-distance bound (l=0.00) with h=0.05 (ablation) | DBG · CIFAR100-LT (ResNet-32, imbalance 100) | overall accuracy: 50.4 -> 50.3 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- DiffuLT — DiffuLT: Diffusion for Long-tail Recognition Without External Knowledge (NeurIPS 2024) · better on 4 of 4 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- DiffuLT: Diffusion for Long-tail Recognition Without External Knowledge (NeurIPS 2024 · similarity 0.84)
- Long-Tailed Learning as Multi-Objective Optimization (AAAI 2024 · similarity 0.84)
- Feature Fusion from Head to Tail for Long-Tailed Visual Recognition (AAAI 2024 · similarity 0.84)
- Generative Data Mining with Longtail-Guided Diffusion (ICML 2025 · similarity 0.84)
- BGDB: Bernoulli-Gaussian Decision Block with Improved Denoising Diffusion Probabilistic Models (AAAI 2025 · similarity 0.83)
- Breaking Long-Tailed Learning Bottlenecks: A Controllable Paradigm with Hypernetwork-Generated Diverse Experts (NeurIPS 2024 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
